We investigated the reliability of the non-invasive approach of measuring steroid hormones from feces in the domestic goose (Anser domesticus), a mainly herbivorous bird with a short gut passage time (2-3 h). Groups of eight outdoor-housed male domestic geese were subjected to three different experiments, injection of either GnRH analogue or ACTH, or "social stimulation" by confrontation with two alien males or females. These experiments were replicated in three different seasons, in spring, during peak reproductive activity, in summer, during long-day photorefractoriness and postnuptial molt, and in fall, during sexual reactivation. GnRH stimulation resulted in significant increases of mean response and peak fecal testosterone metabolites (TM) in spring and fall. Response TM concentrations excreted in spring were generally higher than in summer and fall. Social confrontation with two females, but not with two males, increased mean and peak TM in all seasons. In general, ACTH treatment resulted in a proportionally higher increase of fecal corticosterone metabolites (BM) than GnRH did in fecal TM (80-to 140-fold vs 6-to 8-fold). ACTH significantly increased mean and peak BM in all seasons. Social confrontation with two males significantly increased fecal BM in spring, but confrontation with two females increased fecal BM in fall. Our results indicate that determining steroids from feces is a generally valid approach. However, the sensitivity of the method may vary between different hormones and results may differ between seasons. BM results seemed more sensitive and seasonally robust than did TM.